Evaluation

Evaluation asks whether something worked. The difficulty is defining “worked” before data can answer it and selecting a design proportionate to the claim. rhizae’s current focus is measurement, learning, and evaluation readiness rather than leading high-stakes causal evaluations.

Defining the claim

A useful design starts from the outcome a program claims rather than the one that is easiest to count. Clarifying a vague claim is part of readiness work.

What a design can attribute

A before-and-after comparison shows change; it does not by itself show cause. That distinction must remain visible under the uncertainty standard. A stronger claim requires an appropriate design and, where needed, specialist expertise.

Measuring without adding burden

Indicators that require new collection are weighed against records a team already keeps. Anything collected has to be governed, so each added measure carries the obligations set out in Governance.

Related pilot capabilities are described in Program measurement and learning.